形变攻击潜力
计算机视觉与模式识别
2024-08-05 v1
摘要
在安全系统中,以通用准则测试为意义的风险评估是一个非常重要的课题;这需要依据攻击者的专业水平、其对目标的知识以及设备获取途径来量化攻击潜力。与这些攻击相反,近期揭示的针对人脸识别系统(FRSs)的形变攻击无法由上述任一准则评估。但并非所有形变技术对运行中的人脸识别系统构成相同风险。本文提出形变攻击潜力(MAP)这一一致性方法论,可量化特定形变攻击所创造的风险。
引用
@article{arxiv.2204.13374,
title = {Morphing Attack Potential},
author = {Matteo Ferrara and Annalisa Franco and Davide Maltoni and Christoph Busch},
journal= {arXiv preprint arXiv:2204.13374},
year = {2024}
}
备注
This paper is a preprint of a paper accepted by IEEE International Workshop on Biometrics and Forensics (IWBF 2022). When the final version is published, the copy of record will be available at the IEEE Xplore in proceedings IEEE International Workshop on Biometrics and Forensics (IWBF), Salzburg, Austria, April 2022